Carbon Efficient Tertiary Membrane Bioreactor with Partial Denitrification Anammox for Potable Reuse

INTRODUCTION The Metropolitan Water District of Southern California and Los Angeles County Sanitation Districts (Sanitation Districts) are jointly implementing the Pure Water Southern California (Pure Water) program, which may include building an up to 150 million gallon per day (MGD) advanced water purification facility (AWPF) at the Sanitation Districts' A.K. Warren Water Resource Facility (Warren Facility). The Warren Facility is a 400 MGD high-purity oxygen activated sludge (HPOAS) plant that currently discharges non-nitrified secondary effluent to the ocean ([b]Figure 1[/b]). MOTIVATION Nitrogen management is a key component of Pure Water to meet MWD's nitrogen concentration targets in the purified product water. For the initial phases of Pure Water, the product water will primarily be used for indirect potable reuse (IPR), so the planned membrane bioreactor (MBR) will be configured for nitrification-only operation. For direct potable reuse (DPR) applications, which have lower product water nitrogen targets than IPR applications, the membrane bioreactor (MBR) will need to be configured to support denitrification prior to advanced treatment ([b]Figure 2[/b]) (Sanitation Districts, 2024). To avoid stranded assets when transitioning from IPR to DPR applications, the MBR will include a flexible design (Flex MBR) that allows for multiple operating modes, including using either primary or secondary effluent as the feed water and allowing for future operation to include denitrification. A significant challenge associated with conversion to denitrification is the need for supplemental carbon addition to support conventional nitrification-denitrification (NdN) in the Flex MBR. It is projected that 18,000 gal/day of carbon (MicroC® 2000, EOS, Inc.) will be needed to achieve effluent goals of complete nitrification and ≤ 19 mg NO₃⁻ N/L. To reduce external carbon requirements, two of eight available HPOAS reactors will be operated to achieve NdN. The viability of this innovative process (High-Purity Oxygen Ludzack-Ettinger process or HPOLE) was demonstrated at full scale (Pitt [i]et al[/i]., 2023) and implementation is projected to reduce carbon demand to 14,000 gal/day. Recently, partial denitrification-anammox (PdNA) ([b]Figure 3[/b]) has been demonstrated to significantly reduce carbon demand compared to NdN (Fofana [i]et al[/i]., 2022; Klaus [i]et al[/i]., 2023). Therefore, leveraging knowledge gained from two years of PdNA research at the Warren Facility (Sun [i]et al[/i]., 2024; Farrokh Shad [i]et al[/i]., 2025), the ability to transition Flex MBR process operation to PdNA was included in the full-scale design concept. Modeling projections indicated up to a 60% reduction in demand​, however integration of PdNA with an MBR process had not yet been demonstrated in the industry. RESEARCH OBJECTIVE The objective of this study was to conduct pilot testing to validate the Flex MBR design concept and projected reduction in carbon demand by implementing PdNA. TECHNICAL APPROACH AND KEY FINDINGS The pilot was designed to be a scaled version of the full-scale concept described by Fitzgerald [i]et al[/i]. (2024) and is shown in [b]Figures 4 – 8[/b]. A detailed pilot equipment list is shown in [b]Table 1[/b]. Pilot operation and nitrogen removal performance were continuously monitored with online instrumentation and samples were collected twice per week for laboratory analysis. Anammox activity was confirmed by collecting media from the anoxic zones and batch testing the media to look for indications of annamox activity as described in [b]Figure 9[/b]. In November 2024, the pilot was started-up by seeding with suspended biomass from a demonstration-scale MBR described by Liu [i]et al[/i]. (2025) and operated for one year. At start-up, the system was operated in NdN mode and then virgin media was added to facilitate anammox growth and transition to PdNA. Anammox activity was established after approximately four months of operation. Benchmarked PdNA operating conditions are summarized in [b]Table 2[/b]. Benchmarked performance ([b]Figures 10–12[/b]) demonstrated that effluent water quality goals were consistently achieved ([b]Figure 10[/b]), with average effluent concentrations of <0.1 mg N/L for NH₄⁺ and 11.2 mg N/L for NO₃⁻. Overall, influent total inorganic nitrogen (TIN) loading was reduced by 76% at an observed carbon demand of 2.1 g COD per g TIN removed ([b]Figure 11[/b]), which is approximately 65% lower than the 6 g COD/g TIN reported by Liu [i]et al[/i]. (2025) during NdN testing at the Warren Facility using the same carbon source. Process evaluation indicated that 92% of the TIN removed between ANX 2 and ANX 6 was attributable to partial denitrification–anammox (PdNA), with anoxic NH₄⁺ removal accounting for 21% of the total nitrogen removed across all process tanks, including AER 1–AER 6, highlighting the aeration energy savings potential of PdNA. Anammox activity was further confirmed through ex-situ anoxic NH₄⁺ removal rate testing ([b]Figure 12[/b]). Collectively, these results indicate substantial operational and sustainability benefits for Pure Water, including estimated reductions in supplemental carbon demand from 14,000 to 5,000 gallons per day, annual carbon costs from $20M to $7M, and truck deliveries from approximately 1,400 to 500 per year, corresponding to a 72% reduction in associated greenhouse gas emissions ([b]Figure 13[/b]).